I Sent 1,500 Cold Emails and Got 8 Replies. Here's How I Fixed Our Prospecting Workflow.

It was a Tuesday in February 2024 when my VP of Sales asked if I'd seen the campaign results. I had. 1,500 emails sent. 8 replies. One meeting booked. The meeting turned out to be a recruiter who thought we were hiring.

I manage revenue operations for a B2B SaaS company. I've been doing this for six years and have made—and documented—11 significant mistakes. Together they cost something north of $40,000 in wasted budget and time. This one wasn't the most expensive, but it was the most instructive.

The Mistake I Made With a 'Perfect' Lead List

It started with a simple request: build a list of 1,500 high-intent prospects at companies with 200–2,000 employees. We already had a Wiza LinkedIn email finder Chrome extension installed, so the plan was straightforward. I opened LinkedIn, filtered by title and company size, and pulled the contacts.

I knew I should validate the emails before uploading them to the sequence. But the extension showed a confidence score on every record, and the contacts had come directly from LinkedIn. "What are the odds?" I told myself. Those odds caught up with me fast.

There was also a communication issue that, in hindsight, was entirely my fault. I said "high-intent prospects." The sales team heard "any VP at a company that could theoretically buy us." I never showed them what I meant by intent, because I didn't have a definition yet.

The Fallout Nobody Warned Me About

We uploaded the list, set the sequence to start Wednesday, and by Friday the bounce rate was 18%. Some addresses didn't exist. Some belonged to employees who had left months ago. Some were generic role accounts like info@ and sales@.

The replies that did come in were fewer than my team could count on one hand. One meeting booked, and it was with a recruiter. The whole cycle took about three weeks—or rather, four, counting the recovery.

What most people don't realize is that "verified email" means different things in different tools. A tool can check syntax and domain format and call it verification, but that doesn't tell you whether the mailbox actually exists. A real check needs to go deeper.

I'm not going to claim any tool offers 100% verification accuracy, because none do. But Wiza does real mailbox-level verification, and even then we keep a small buffer for the addresses that die the day after validation.

The New Workflow: Validation First, Volume Second

The fix was process, not magic. I changed the workflow and now use these steps for every campaign:

• Every list gets run through an email validation API before it goes near the CRM.
• Any campaign over 500 records gets a second pass with an email validation service.
• Role-based addresses get filtered out unless we intentionally target a function like sales@ or support@.

According to Gartner's 2024 B2B Buying Study (gartner.com), 75% of B2B buyers now pick their own buying process, and much of it happens before a seller is invited. The email either earns a reply or it doesn't. Validation doesn't make your message better—it makes sure your message is actually read.

Wiza Pricing: When the Yearly Plan With Intent Topics Makes Sense

I can't tell you whether Wiza is right for every team. I can tell you why we chose the yearly plan with intent topics.

After the disaster, I stopped pulling lists by job title alone. Intent topics data tells us which accounts are actively researching things like "sales engagement tools" or "revenue intelligence." That's not the same as who has the right title. It's a much more useful signal.

Wiza pricing on the yearly plan felt painful at first. But the cost of the failed campaign, including the hours my team spent scrubbing the list and repairing our sender reputation, was far higher than the price of the data we should have started with.

So How Does AI Cold Email Fit into an Agent-Native Prospecting Workflow?

The phrase "agent-native" gets thrown around a lot. For us, it means AI agents own the repetitive parts: enrichment, scoring, segmentation, and drafting. Humans own judgment: reviewing the shortlist, editing the language, and deciding who gets sent what.

AI cold email fits into that workflow after the data is clean. Wiza's AI email writer, for example, takes a prospect's industry and intent topic and produces a short personalized opening. It's a starting point, not a finished message. Our reps edit it, add specific context, and then send.

People think AI-generated email is what drives replies. Actually, relevant and carefully sent email drives replies. AI just makes the "carefully" part scalable. If you feed an AI bad data, it writes beautiful emails that bounce. If you give it no intent signal, it writes generic emails that sound like spam.

The Result and the Real Lesson

After we implemented the validation step and intent topic filtering, the next campaign looked completely different. We sent 1,200 emails with a bounce rate below 2%. The team got 27 replies and 9 qualified meetings. Those aren't earth-shattering numbers, but compared with 8 replies and a recruiter, it felt like a different universe.

I'm not going to tell you Wiza is the only tool you'll ever need. That kind of universal claim is a red flag. For some companies, a global data provider or a dedicated intent platform will make more sense. But the principle applies anywhere: check the data, run the validation, keep a human in the loop, and know exactly what your tool is good at.

The lesson I teach our new hires is simple: check the data before you spend the money, and keep a human in the loop before you hit send. Cold email still works. It just doesn't work on bad data.